Beginner: Motorised Camera Pan and Tilt

Hi,

I’m new at this, and need some help getting started on my project. I have done a lot of programming before, but nothing like this.

I would like to build the following:

A small unit that can pan and tilt my camcorder (350 grams).
It should fit on top of my tripod and fit in a 1.0" pipe. ( I know… tiny)
I should be able to have wireless control (via remote control or wifi or Bluetooth ect)
It should be able to program it. ( Eksampel: if I press “1” on the remote control then turn left 5° at x speed and 3° up at y speed)
Run on batteries

How do I get started? Can I use the “Baby Orangutan B-168 robot controller” as my core? What motor(s) should I choose?

Thank you for any and all help!

Sounds neat, and it would be easy to do with hobby servo motors if you could relax one of your requirements, either the camera weight or the pipe diameter. I’m guessing your camera isn’t going to magically get any lighter, but servo motors that can fit in a 1" pipe aren’t very powerful.

There are lots of different ways to go here though, so it’s time to answer your question with a bunch of mine:

-Does the motorized pan-tilt mechanism have to fit in a 1" pipe, or just the electronics? After all, the 350 gram camera can’t fit in a pipe anyway (or can it?).

-Do you want absolute position control (i.e. turn to 30 degrees from straight ahead and 25 degrees up from horizontal) or just relative position control (i.e. turn left a little and up a little)?

-Do you really need speed control, or just to be able to get where you want it to point?

-What sort of budget are you planning on?

And just out of general curiosity, could you tell us a little more about your project?

If you don’t mind having the mechanism outside of the pipe, Servo City makes a nice little pan-tilt module here. Depending on how you decide to go, you can control a couple of servos with a Baby Orangutan (the slightly cheaper B-48 would probably do just fine, but the B-168 gives you lots of room to expand). You’ll also need a programmer to go with the microcontroller, but there are some combo deals. You could also use just a simple Micro Serial Servo Controller and a serial radio of some sort, with all the intelligence on the remote control side. Either of these would fit in a 1" diameter pipe.

-Adam

P.S. If you do use hobby servos and you’re willing to settle for direct manual control you can just buy a hobby transmitter/receiver pair like the kind made for high-end RC cars and planes to control the servos with the controller’s joystick:

Hi

So I can keep on top when i use it as a Hiking Staff. If it’s bigger than 1" pipe I could take it off and put it my pocket i guess…

Yes, there are some devices that does relative position control. (Ebay $50). I need full control to get the range of automatic shots I’m looking for.

Yes, I need speed control. Sometimes filming a slow pan of the landscape and sometimes just change camera angel as fast.

well, $100 - $800 ??

I do adventure travel videos on my own. I would like to be able to do a range of automatic shots (with or without myself in the frame.) Smooth shots is not easy to make in the field. It would add a lot of production value to my projects.

It looks like something i might use. It is a little big … But thanks! :smiley:

More ideas?

Aah, I see. I think one of the Servo City pan-tilt platforms would work very nicely for that. The only downside to using servos to directly drive your pan-tilt platform is that you only get 180 degrees of motion on each axis. If you want compact absolute position control, a hobby servo platform is by far the simplest way to go.

What you need in addition to the motorized pan/tilt platform is a way to control the servos, and two ways to do this jump out at me.

If you want the flexibility to write new motion sequences while you’re out in the field, you could bring a laptop with a serial radio, and have it send commands to a receiver, which you could connect to a micro serial servo controller. The Pololu servo control protocol has a speed control option, so you would essentially be commanding new positions to the servo motors and how fast to travel there.

If you’re okay with picking from a set of pre-programmed motion sequences, you could program a Baby Orangutan to generate the servo signals directly (it’s nice to have a dedicated controller for lots of servos, since the signals are all about timing, but its pretty trivial for one or two servos). There are a couple of ways you could remote control the Orangutan (serial radios, with a laptop or another microcontroller on the other end), but if you’re just picking from one of several preprogrammed sequences you also just wire up a couple of buttons to select from these sequences, and a long delay to let you get out of the way.

For the couple of extra bucks, I would go with one of the ATMega168 based controllers to have ample room to store lots and lots of motion sequences. A Baby Orangutan will fit in a 1" diameter pipe, but you might even want to go with a big Orangutan, which has an LCD screen and built-in buttons, so you could set it up to scroll through named sequences and tell you what you’re doing. Both of these are available as combos with programmers.

Any of this sound good?

-Adam

P.S. Have you taken a look at Phil’s Panobot thread? He built a larger pan/tilt rig for taking panoramic photos with a still camera, and it turned out very well (I know of at least two people who are duplicating the project).

Do you know how I figure how what kind of torque i need on the servos?

Sure, lets do some “back of the envelope” calculations (as in rough and simple enough to be scribbled on whatever paper is handy).

Your camcorder weighs 350 grams, or about 12.35oz. If it’s like my camcorder it’s about four inches tall, and we can assume the center of mass is in the center of the camera, so two inches up off of the platform the camera rests on. The platform of the SPT200 Pan & Tilt System is about an inch above the rotation axis of the tilt servo (the one that will actually be working hard), so let’s say you have a 12.35oz mass on a 3inch lever arm. In the worst case scenario, when the platform is rotated 90 degrees from horizontal, the motor is going to have to generate a minimum of 12.35oz*3inch~=37onuce-inches of torque just to keep the camera still. It really is that simple.

To actually move through this angle and not just stall the motor, you’ll want a servo that can deliver at least twice that torque. If you’re not going to be looking straight down much, the HS-475HB will do just fine. If you are going to go to extreme angles like straight up and down a lot, and you want to insure smooth motion, three times the minimum torque will be plenty, and the (twice as expensive, but still under $40) HS-645MG would be a good choice. It’s nice to have two of the same kind of servo (you don’t have to worry about which is which, you can measure just one signal-angle ratio, etc…) but there is no reason that the pan servo has to be this strong, so a HS-475HB will do just fine.

By the way, if you do buy servos from Servo City, you can totally skip the $10, 180 degree rotation modification option. Pololu servo controllers, or servo pulses you generate with your own microcontroller, can drive these servos to their ~180 degree extremes with no modification.

-Adam

I don’t know what happened to my earlier post. Will repeat again.

My son is an avid basketball player. I took on the task of recording his games so we can review it afterwards. I have a digital camcorder mounted on a tripod and all I have to do is to pan the camcorder horizontally to record. Unfortunately, I kept forgetting to rotate the camcorder because I am too busy watching the game. I’ve lost many great shots or moves because of that and it’s very frustrating. If I can only turn back the clock and record again.

After reading this thread, it gives me hope. I like to (somehow) mount a small transmitter on him and have the camcorder track him on the court - just a simple horizontal pan of, say 90 degree should be sufficient (place the tripod on one end of the court and pan to record action of the whole field). However, what wireless should I choose? Bluetooth wont work probably (too far). RF is most likely too unreliable. What do you guys think?

Well, this is exactly what I need:

21best.com/21_best/electroni … anual.html

but for a cool $289 bucks!!!

Besides, I like to build things myself…now where do I start?

Mmmmm…that one has the right idea but its only good for a 35’ circle (uses ultrasound) and the Senor is too clumsy . May be I should go for a computer controlled pan and tilt and uses a separate RF interface - all under software control. Would I be able to use a small transmitter mounted on the player then?

Hi,

This sounds like a hard project. What if you start with just two camera positions, one for when either team has possession. You can probably detect possession easier than where your son is.

- Ryan

Yes but where is the fun?

Unless you mean have two cameras mounted and have a program compares inputs and move them accordingly?

If you mean having one camcorder pointing to each end, that would be horribe because the resulting video would be pretty bad. Besides, you loose all the transitional plays.

The link I cited has the right idea, if not because of the 35’ distance limitation

Hello.

As Ryan has pointed out, this sounds like a very complicated project, so I don’t think you’re going to get much advice here beyond perhaps some very high-level suggestions. It would be helpful to know your level of experience with projects like this. For example, do you have much programming experience with computers or microcontrollers? How comfortable are you with electronics?

- Ben

Programming and electronics is no issue (MSEE). Just want to see what’s available out there and don’t like spending hundreds for it. Love to get my hands dirty. I don’t get to do this kind of stuff since college. Always wanted an excuse to build something fun.

Upon further research, some guy sells a quad-camcorder pan-and-tilt with Windows drivers for it (cgi.ebay.com/4-Camera-Wireless-C … 23100c3378). So, now if I can figure out which sensor to use - or by-pass the sensor and use software to determine which of the 4 camcorders is recording procession…